A dynamic ensemble pruning framework (DEPF) is proposed to identify and interpret single-cell molecular heterogeneity. In particular, a silhouette coefficient-based indicator is developed and evaluated to determine the optimization direction of the bi-objective function. Following that, a bi-objective fruit fly optimization algorithm is designed to prune dynamically the low-quality basic clustering in the ensemble.
H. Zhu, Y. Yang, Y. Wang, F. Wang, Y. Huang, Y. Chang, K. Wong*, X. Li*
A dynamic ensemble pruning framework (DEPF) is proposed to identify and interpret single-cell molecular heterogeneity. In particular, a silhouette coefficient-based indicator is developed and evaluated to determine the optimization direction of the bi-objective function. Following that, a bi-objective fruit fly optimization algorithm is designed to prune dynamically the low-quality basic clustering in the ensemble.
Z. Yu, Y. Su, Y. Lu, F. Wang, S. Zhang, Y. Chang, K. Wong*, X. Li*
A dynamic ensemble pruning framework (DEPF) is proposed to identify and interpret single-cell molecular heterogeneity. In particular, a silhouette coefficient-based indicator is developed and evaluated to determine the optimization direction of the bi-objective function. Following that, a bi-objective fruit fly optimization algorithm is designed to prune dynamically the low-quality basic clustering in the ensemble.
Y. Fan, Y. Wang, F. Wang, L. Huang, Y. Yang, K. Wong, X. Li*
A dynamic ensemble pruning framework (DEPF) is proposed to identify and interpret single-cell molecular heterogeneity. In particular, a silhouette coefficient-based indicator is developed and evaluated to determine the optimization direction of the bi-objective function. Following that, a bi-objective fruit fly optimization algorithm is designed to prune dynamically the low-quality basic clustering in the ensemble.
Z. Zheng, J. Chen, X. Chen, L. Huang, W. Xie, Q. Lin, X. Li*, K. Wong*
A dynamic ensemble pruning framework (DEPF) is proposed to identify and interpret single-cell molecular heterogeneity. In particular, a silhouette coefficient-based indicator is developed and evaluated to determine the optimization direction of the bi-objective function. Following that, a bi-objective fruit fly optimization algorithm is designed to prune dynamically the low-quality basic clustering in the ensemble.
F. Wang, H. Alinejad-Rokny, J. Lin, T. Gao, X. Chen, L. Meng, X. Li*, K. Wong*